The Economic Burden of High School Dropouts and School

0

The Economic Burden of High School Dropouts

and School Suspensions in Florida

Clive R. Belfield Queens College

City University of New York

Center for Benefit-Cost Studies in Education Teachers College, Columbia University

November 2014

Commissioned by The Civil Rights Project/Proyecto Derechos Civiles at UCLA

with support from The Atlantic Philanthropies

The author appreciates comments from Daniel Losen and Russell Rumberger.

1

The Economic Burden of High School Dropouts and School Suspensions in Florida Center for Civil Rights Remedies, Civil Rights Project/Proyecto Derechos Civiles, Nov. 2014

Summary In this paper, we calculate two economic consequences of Florida public school students dropping out of high school—the social consequences for the state economy and well being, and the fiscal consequences for the federal, state, and local governments. Approximately one in five Florida students—more than 50,000 in each age cohort—will not graduate from high school. Over their lifetimes, these dropouts will have lower incomes, contribute less in taxes, rely more heavily on government health and welfare programs, and impose higher costs on the criminal justice system. Using evidence from research studies and Florida-specific databases, we model the social and fiscal gains that would accrue if more dropouts became high school graduates. The model yields separate results by sex and race. The social gains—which include earnings, health, crime, welfare, and productivity spillovers—are calculated at $354,000-$476,000 for each additional high school graduate. This amount is the value at the time of graduation. The fiscal gains for Florida are calculated separately for the state/local government and the federal government. The direct state/local fiscal gain per additional high school graduate is $57,000-$63,000, whereas the fiscal impact on federal spending in the state is $59,000-$89,000. The full fiscal gain per new high school graduate is therefore $116,000-$152,000. The model is based on conservative parameter values and assumptions regarding the benefits of educational attainment. Based on sensitivity testing, we show that, even making highly restrictive assumptions, the social and fiscal gains are economically meaningful. Evidence shows that school suspensions adversely impact high school graduation rates and that the dropout rate would be much lower if there were fewer suspensions. Using a new economic model, we can calculate the economic losses under different scenarios.

2


Table S1: Aggregate Burden of High School Dropout and Suspension Policies for Florida

Status quo

policy

Total impact of suspensions on HS graduation

If suspension rate were

reduced by 5 percentage

points Suspension rate 27% 27% 22% Suspended students per cohort 48,850 48,850 39,750 Reduction in high school dropouts -- 3,220 600 Social Loss averted ($ m) -- $1,531 $285 Fiscal Loss averted ($ m) -- $488 $91

Across the student population in Florida, 27% students in each ninth-grade cohort will be suspended. Current school suspension policies and practices reduced high school graduation rates by six percentage points, resulting in 3,220 fewer high school graduates. This produces a total social loss of $1.5 billion and a total fiscal loss of $0.5 billion for each cohort of ninth-grade students. Schools and districts could simply reduce their high school suspension rates, which would have an indirect effect on the dropout rate. For example, if the suspension rate fell by five percentage points (from 27% to 22%), the number of high school dropouts would fall by 600. This would avert social losses of $285 million and fiscal losses of $91 million.

3


Contents

1. Introduction ................................................................................................................. 4

2. Educational Attainment in Florida ............................................................................ 5

3. Modeling the Economic Value of Educational Attainment ..................................... 7

4. Economic Benefits by Attainment Level .................................................................. 8

4.1 Earnings and Tax Payments .................................................................................... 8

4.2 Other Fiscal and Social Benefits ........................................................................... 11

5. The Economic Burden of High School Failure in Florida ..................................... 15

5.1 Lifetime Burdens of High School Dropouts ......................................................... 15

5.2 Sensitivity Analysis ................................................................................................ 17

5.3 The Economic Burden of High School Suspensions .......................................... 19

6. Conclusions .............................................................................................................. 20

References .................................................................................................................... 20

Appendices ................................................................................................................... 35

4


1. Introduction

There is an immense amount of evidence on the economic and social value of

educational attainment (Belfield and Levin, 2007; Lochner, 2011; Oreopoulos and

Salvanes, 2011). The impact education has on earnings is well established. There is

also extensive literature on the health advantages of having more education, and

evidence on the social benefits of having a better educated population. One important

and often overlooked benefit is that the gains from education are greater for minority

and disadvantaged students; thus, education has the potential to narrow gaps in

economic well-being (Hoxby and Turner, 2012). Many youth without a high school

diploma will face a precarious economic future (Fernandes-Alcantara, 2012).

Despite this evidence, many students do not complete high school. Nationally,

graduation rates have remained static over recent decades, with one-fifth to one-quarter

of students failing to graduate on time. Moreover, high school graduation rates are

much lower for minority students and low-income students, and for male students

(Knapp et al., 2011, Table 7). Many high school graduates do not go to college—only

two-fifths of 18-24-year-olds are in college—and many college students fail to complete

a degree: the U.S. college graduation rate for four-year college students is just over

one-half; at two-year colleges, the graduation rate is one-third (NCES Digest, 2012,

Table 213). As with high school, college enrollment and completion rates are

significantly lower for minority students and low-income students. These low rates are

partly a result of students not being adequately prepared for college when they finish

high school.

Hence, there is a strong economic case for increased investments to improve

school quality and to ensure that students graduate from high school. This is especially

true for disadvantaged students, many of many of whom are dealing with family

disadvantages and other barriers that undermine their ability to complete high school.

For these students, school policies and practices can be especially influential,

One such policy is school suspensions. Schools can and do suspend students

for a number of reasons, from criminal behavior, such as bringing a weapon to school,

to more minor offenses, such as swearing or “willful defiance” of teachers or

5


administrators (Fabelo et al., 2011). School personnel have a lot of discretion in terms

of how to manage student behavior, and they do have options preferable to out-of-

school suspensions, such as in-school suspensions or mandatory intervention

programs, such as Restorative Justice.1 Nevertheless, many students end up under the

jurisdiction of the Department for Juvenile Justice.2

In this paper, we calculate the economic consequences of high school

suspensions in Florida. We do this by first estimating the economic losses resulting from

high school dropouts in Florida and then by estimating the impact suspensions have on

those dropouts. It is important to understand the full economic consequences of

students failing to complete high school. There are fiscal consequences: tax revenues

are lower and government spending is higher for dropouts; and there are social

consequences: the state economy is adversely impacted by dropouts’ lower productivity

and other negative factors, such as crime and poor health. We calculate both the fiscal

and social impacts using a conventional lifecycle economic model. The initial step is to

map the distribution of educational attainment across cohorts of Florida youth and

government spending across the state. We then describe the economic model and how

key components of the model are derived. The results of the model are expressed as

the net gain to Florida taxpayers and Florida residents if high school dropouts instead

completed high school. It is then possible to simulate the gains that would accrue if

education policies and high school suspension practices were more effective and the

high school completion rate in Florida were higher.

2. Educational Attainment in Florida

Each age grade in the Florida public school system includes approximately 210,000

children; 41% of the students are white, 30% Hispanic/Latino, 23% black, and 6% other

race/ethnicity.3 However, by 12th grade, only 160,000 of those students remain in school

and complete their high school diploma (including certificates but not GED). Accounting

for mobility, the high school completion rate for the Florida public school system is 76%,

1 See restorativejustice.org. 2 See http://www.djj.state.fl.us/docs/research2/2010-11-delinquency-in-schools-analysis.pdf?sfvrsn=0. 3 All education data is from www.fldoe.org/eias/eiaspubs/pubstudent.asp; www.fldoe.org/fcs/colleges.asp and www.flbog.edu/resources/ditr/.

6


which means that one-in-four students does not graduate on time. This rate is close to

the national average.4

There are important racial differences in high school completion rates in Florida,

which are significantly lower for minority students: the rate is 88% for whites, 72% for

other racial groups, 71% for blacks, and 65% for Hispanic students. There also are

significant gender differences: the graduation rate for female students is five percentage

points higher than for male students. These rates are similar to the differences found

nationally.

High school attainment in large part drives college enrollment rates. One-third of

all Florida high school graduates will attend a two-year Florida public college, and one-

quarter will attend a four-year Florida public college. (One in ten college-bound students

will either attend a private college or go to college out of state). Again, there are

significant differences by racial group and by gender.

The distribution of educational attainment for the most recent cohort of high

school graduates in Florida is given in Figures 1F and 1M. Overall, of the 210,000

students in each Florida cohort, 52,900 will not graduate from high school. This group is

disproportionately male and minority. Of those who do graduate, the majority (98,200)

will not go on to attend a postsecondary school; one-third will attend college, but

approximately half of those will not complete a four-year degree (34,900); the remaining

half will complete at least a bachelor’s degree (34,600).

Many factors influence educational attainment, in particular the large number of

high school dropouts. One of the most influential factors is likely school practices, which

also may be the most malleable from a policy perspective (Rotermund, 2008).

The focus here is on school suspension practices. As described in detail in Losen

and Gillespie (2012) and Losen and Martinez (2013), millions of K-12 students are

suspended from school for some time each school year; the suspension rates have

grown dramatically over the last two decades. The incidence is much higher for black

students: one in six black students is suspended from school at least once during the

4 The state’s average freshman graduation rate, at 71%, is actually below the national average of 78% (Stillwell and Sable, 2013, Table 1). See also Chapman et al. (2011).

7


school year, which is more than double the average rate.5 Importantly, suspension

policies are often at the discretion of school personnel and are not determined simply by

student behaviors; suspension rates therefore differ substantially across schools,

districts, and states (Fabelo et al., 2011).

Aggregate data on suspension rates within the Florida K-12 system are difficult to

estimate, but the most recent estimate is that 27% of students are suspended while in

high school (Losen and Gillespie, 2012). Student-level analysis shows that suspension

has a very adverse effect on high school completion (Fabelo et al., 2011). Based on

evidence from Balfanz et al. (2014), ninth-grade students in Florida who are suspended

graduate from high school at rates that are 6.6-7.3 percentage points lower than

students who are not suspended. Therefore, if school suspension practices were

changed—through enhanced resources to avoid the need for suspension and/or more

effective supports for students who are suspended—high school graduation rates

should increase significantly. We calculate the economic gains higher graduation rates

would generate for the state of Florida and for its taxpayers.6

3. Modeling the Economic Value of Educational Attainment

The basic framework for an economic model to value educational attainment is

described in detail in Belfield and Levin (2007). There have been numerous applications

of this model at the national, state, and local level, as well as for subgroups of youth.7

Thus far, the model has not been applied at the state level in Florida.

The details of the model are given in Appendix Box 1. The model operates

through differences in educational attainment at the high school level, which then have

implications for changes in educational attainment at the postsecondary level. Students

who fail to complete high school are classed as dropouts. Students whose education

ends with high school are classed as high school graduates. The third group are 5 Disciplinary referrals to law enforcement vary significantly across racial groups: minority students represent almost 60% of all students in Florida, but they account for 66%-69% of all law enforcement referrals within the school system (Civil Rights Data, 2009). 6 From this economic perspective, it does not matter what has caused the suspension (Wright et al., 2014). Also, the model does not include any spillover effects on other students (Kinsler, 2011). At issue are the economic consequences of suspending each student from school. 7 For example, Trostel (2009); Sum et al. (2009); Baum et al. (2010); Gottlob (2007); Carroll and Erkut (2009); Belfield and Levin (2009); Bush School of Government and Public Service (2009).

8


“expected high school graduates,” who follow the typical path of a low-income high

school graduate—that is, one-third go on to college and one-sixth complete a bachelor’s

degree.8 As the distribution of students in each group changes, there are private, fiscal,

and social consequences, which include changes in earnings and productivity, as well

as other changes that are translated into monetary terms. The focus of this analysis is

the separate fiscal consequences for state/local government in Florida, and the

aggregate social consequences for the state.

The economic metrics created from the model are the individual and aggregate

present value of benefits associated with different levels of educational attainment. All

economic calculations are based on the most recent social science evidence

disaggregated by sex and race (white, black, Hispanic, and other racial groups). Where

available, educational and labor market data for Florida are used; government

expenditure data are taken from Florida state departments.

All money amounts are expressed in present values for students at age 18, using

a discount rate of 3.5% (Moore et al., 2013). All figures are in 2014 dollars, weighted to

the prices in Florida. The dollar amounts generated by the model represent lump sum

values when a student is age 18. These values can therefore be considered equivalent

to a bank deposit made at that date.

4. Economic Benefits by Attainment Level

4.1 Earnings and Tax Payments

More education leads to higher lifetime earnings; the association holds strongly, even

accounting for differences in student ability and motivation. Consequently, more

educated people contribute more in federal, state, and local taxes. These effects hold

across all levels of education, and there is some evidence that they are especially large

for minority students.9

For this model, lifetime earnings and tax payment profiles are calculated for four

pathways: high school dropouts, high school graduates, people with some college, and 8 Gender- and race-specific college progression and completion rates for Florida students are applied in the model. 9 Belfield and Bailey (2011); Altonji et al. (2012); Avery and Turner (2012); and Dale and Krueger (2011).

9


those with a four-year degree. The last three pathways are then combined to create the

pathway for an “expected high school graduate,” based on rates of college enrollment

and completion. Separate profiles are created by gender and racial groupings. The

pathways can then be compared in terms of the differences in lifetime earnings and tax

payments.

Two data sources are used for earnings and the average profile is applied. The

first dataset is merged March Supplements of the Current Population Survey (CPS) for

the years 2009-2013. The CPS includes data on gross earnings for 40,530 Florida

residents age 18-65. The second dataset is the Public Use Micro Sample of the

American Community Survey (ACS) for the five-year period 2006-2010. The ACS

includes data on personal incomes for 930,180 Florida adults. Although the CPS is

devised to calculate earnings with the most precision, the ACS has the advantage of

larger samples for subgroup analysis within a single state.10

Earnings profiles reflect gross earnings plus health benefits, adjusted for labor

force participation rates and productivity growth.11 The profiles of groups with more

education are also adjusted for ability/motivation.12 For each dataset, earnings are

collapsed into education levels and five-year age bands. From these age bands, each

full lifetime earnings profile is extrapolated to age 65 and then discounted back to a

present value at age 18.

Table 1 shows the present-value lifetime earnings for a Florida high school

student by sex and racial group. Male [female] high school dropouts are predicted to

earn $406,430 [$206,750]. By contrast, high school graduates are predicted to earn

$623,770 [$397,410], a gain of approximately $200,000 over dropping out. An expected

high school graduate (where the probability of going to college is accounted for) will

earn $760,110 [$493,960]. This amounts to a gain of $353,680 [$287,210] over a

10 Results from the CPS and ACS are within +/-10% for each gender-race profile. 11 These benefits also vary with education levels. Based on the CPS, workplace health insurance rates for female [male] workers are 11% [19%] for dropouts, 29% [39%] for HS graduates, and 39% [51%] for the expected HS group (CPS data). Based on the Medical Expenditure Panel Survey, private-sector health insurance plans are $5,340 per annum; see http://meps.ahrq.gov/mepsweb/data_stats/summ_tables/~ insr/national/series_1/2012/tic1.htm. 12 Typically, studies find that adjusting for ability and student motivation has little effect on the gains to college (Carneiro et al., 2011; Rouse, 2007). Conservatively, a downward adjustment (alpha factor) of 0.1 is applied.

10


Florida high school dropout. Although the gaps by education level differ by sex and

race, they are all substantial.13

A separate approach is followed for tax payments. Federal incomes taxes are

calculated in three approaches and then averaged. The first approach uses declared

after-tax federal income tax payments made by respondents to the CPS (adjusted for

labor force participation, productivity growth, and ability). The second approach uses

gross earnings data (derived as above from the CPS and ACS) and applies the NBER

TAXSIM calculator. Finally, federal income taxes are calculated as a flat proportion of

gross earnings based on simple marginal tax rates.14

State/local tax payments are calculated in a similar way. Florida has no state

income tax; state tax revenues come primarily from sales taxes, with property taxes

levied at the county level (see Appendix Table 1). State taxes are calculated in two

ways. The first uses the CPS data to derived after-tax state tax payments and property

tax payments. The second way is to calculate state sales taxes as a proportion of gross

earnings (from CPS and ACS data); the tax proportions are based on current state and

county sales tax rates (adjusted for exemptions). The results from these approaches are

averaged to create lifetime state/local tax profiles.

Table 2 shows present value tax payments by education level and level of

government by sex and race. Tax payments are increasing with each education level. At

the state level, a high school dropout will pay in $49,700 [$45,430], a high school

graduate $72,800 [$61,350], and an expected graduate $82,650 [$70,090]. The net

effects of education are payments approximately $20,000-$30,000 higher than those of

dropouts. Again, these education gaps are present across all groups. However, they are

somewhat lower than in other states because of the Florida tax structure and, notably,

the absence (along with six other states) of a state income tax.

In addition to state/local taxes, Florida residents pay federal incomes taxes,

which are increasing with education levels. The present-value lifetime profiles of federal

13 Accounting for ability adjustments and Florida-specific prices, these estimates accord with national estimates by Oreopoulos and Petronijevic (2013) and Avery and Turner (2012). 14 Based on 2013 marginal tax rates, incomes above $40,000 are assumed taxed at 25% and incomes below $40,000 at 15% (www.irs.gov). FICA contributions are excluded.

11


income taxes are given in the bottom panel of Table 2. Relative to a high school

dropout, each high school graduate contributes $41,290 [$28,370] more in federal taxes

and each expected high school graduate contributes $70,240 [$47,410].

4.2 Other Fiscal and Social Benefits

Education conveys additional benefits over the lifetime, which has both fiscal and social

impacts. The fiscal impacts are divided according to whether the financing is at the

state/local or federal government level, shown in Table 3. The social benefits are shown

in Table 4.

Educational attainment is strongly positively associated with healthier behaviors

and thus with better health. For example, college graduates exercise more regularly and

report lower rates of diabetes and obesity.15 Individuals with more education also have

jobs that provide private health insurance. These health behaviors and health status

associations hold even after controlling for income, and they are found across all

increments of education and almost all health conditions. They also hold across years of

high school and extra years of college education.

Improved health status is both socially valuable—it is desirable to have healthy

citizens—and fiscally valuable—government spending on health is lower for a higher

level of health. Differences in health status affect government health expenditures.

The social value of improved health is calculated using Quality Adjusted Life

Years (QALYs) —that is, years of life relative to years in full health. Schoeni et al.

(2011) estimate annual differences in health-related quality of life across individuals with

different education levels: conservatively, each year of education adds 0.008 QALYs

annually during adulthood.16 The value of a QALY is assumed at $100,000 (Cutler and

Lleras-Muney, 2010). Hence, each year of education yields a “health annuity” of at least

$800. Extrapolated over a lifetime, this represents a significant social benefit. 15 On obesity, see Rosenblum (2012). On education-health gradients, see Cutler and Lleras-Muney (2010), Kimbro et al. (2008), Oreopoulos and Salvanes (2011). On the health effect of selective colleges, see Fletcher and Frisvold (2011). 16 An alternative study yields higher estimates of the social value of health. Using MEPS and NHIS data from 1997-2002, Muennig et al. (2010) calculated the remaining QALYs of persons reaching adulthood. At age 16, persons with more education have many more QALYs ahead of them: high school graduates are expected to have 1.5-2.4 more QALYs over their lifetime than dropouts. To be conservative, this study is not applied in the economic model.

12


The fiscal health burden across education levels is calculated as the average of

two methods, using national surveys. The first method is based on Medicaid/Medicare

enrollment rates by education level from the ACS; these are multiplied by average per-

enrollee spending in Florida through adulthood (Sum et al., 2009). The second method

is based on national estimates that include Medicaid/Medicare and Social Security

Disability payments up to age 65 plus direct state-level medical spending (Muennig,

2007).

Calculated in this way, these social and fiscal health gains are conservatively

estimated. They do not account for the steepening health-education gradient—that is,

the fact that education is becoming a more important determinant of health (Adler and

Stewart, 2010). Also, no adjustment is made for increases in healthcare costs, which

are anticipated to exceed inflation (in part because of the Affordable Care Act).

Criminal activity is also strongly correlated with education levels (Lochner, 2011;

Lochner and Moretti, 2004; Oreopoulos and Salvanes, 2011). There is a direct

behavioral association—for a variety of reasons, more education may reduce criminal

activity—and an indirect income effect—education raises incomes and so increases the

opportunity cost of crime.17 Lower crime is a social good, leading to improvements in

quality of life and fewer burdens on victims. It is also fiscally valuable: where crime is

lower, spending on policing, the criminal justice system, and incarceration is reduced.

The benefits education has on criminal activity are derived in two steps. First, the

lifetime fiscal and social crime burden is calculated per non-offender, per general

offender, and per chronic offender. Second, the probabilities of being in each category

are modeled according to education levels.18

17 As shown in Appendix Table 2, Florida is not a low-crime state; overall, the violent crime rate in Florida is higher than the national rate by more than one-quarter (26%). Specifically, rates of aggravated assault, burglary, and larceny theft are all significantly higher than the national average. Also, the property crime rate exceeds the national average by 15%. 18 Non-offenders (about 80% of the population) impose zero crime burden. General offenders (15%) commit about half of all crimes. Chronic offenders (5% of the population) commit the remaining half of all crimes (Farrington and Welsh, 2007). Hence, chronic offenders impose the largest burden. The lifetime fiscal and social consequences of being either a general or chronic offender are taken from DeLisi et al. (2010) and Cohen and Piquero (2009), adjusted for federal and state rates of probation, parole, and incarceration (SCCJSO, 2012)). The social burden of crime can also be estimated as a proportion of the fiscal burden (Miller et al., 1996). The relative crime probabilities are from Lochner and Moretti (2004).

13


These crime estimates do not count some important costs for which accurate

data are not yet available. Specifically, the psychic costs to victims’ families are

unknown, as are the long-term effects of being a victim of crime; also, no reliable

estimates exist of the loss in property values (and property taxes) for high-crime

neighborhoods. As with health, more recent calculations suggest that the burden per

crime is increasing faster than inflation.

There are many other economic benefits from educational attainment. Three of

these are included, as there is sufficient evidence to estimate their monetary value.

First, higher levels of education are associated with lower welfare expenditures

(partly because welfare is means-tested). In Florida, more than 90% of adult TANF

recipients are high school dropouts (DHHS, 2012, Table 10:25); for high school

graduates and college graduates, respectively, housing assistance gaps are 1% and

35%, while food stamps gaps are 19% and 54% (Barrett and Poikolainen, 2006;

Waldfogel et al., 2007). Based on caseload expenditures and these education-welfare

gradients, significant fiscal savings result from higher education levels.19 These fiscal

savings are split between the federal government and state/local governments as a

function of their relative burden of support across all welfare programs. From the state’s

perspective, however, federal expenditures are tied to maintenance-of-effort

agreements: savings to the federal government are therefore also valuable for state

governments. Fiscal savings are calculated as the total expenditures on social services

programs, but from a social perspective, welfare spending is a transfer from taxpayers

to welfare recipients and so has neutral resource effects (albeit distributional

implications). Social savings are therefore counted as the administrative resources

needed to effect welfare transfers.

A second benefit is from labor productivity spillovers (Monaco and Yamarik,

2013). Educated workers are more productive when working with other skilled workers,

in part because they can learn from each other. As the proportion of college graduates

in a population increases, so do average earnings and gross state product. These 19 Rates of welfare reliance and amounts are from the American Community Survey (Tables ACSBR/09-13 and ACSBR/09-8) and from the Department of Health and Human Services, Administration for Children and Families (Table_b1_2009.htm, 2012, 10:25). The economic burden of welfare includes TANF, SNAP, housing vouchers, and state welfare programs (excluding Social Security payments).

14


spillovers are conservatively estimated at 6% of individual earnings (Abel et al., 2010;

McMahon, 2006).

The third benefit accrues from savings to tax collection and from reductions in the

distortionary effect of taxation.20 This “marginal excess tax burden” (METB) is calculated

for all fiscal savings (and costs) arising from higher educational attainment. The METB

has been found to be substantial. A conservative estimate is 13%—that is, for each

dollar saved in expenditure, the full social gain is at least $1.13 (Allgood and Snow,

1998). The rate is probably higher for state/local taxes, which are imposed on goods

with inelastic demand. From a fiscal perspective, the METB is important because taxes

would be collected on the distorted economic activity.

Finally, the additional costs of college education are included. Attending college

generates additional costs along with the benefits itemized above, and these costs

should be included in the cost-benefit calculations. College costs—tuition and

government subsidies—are calculated separately for two-year colleges, four-year public

colleges, and four-year private colleges in Florida. The costs are then apportioned

across students in the expected high school graduate group, based on their respective

enrollment patterns in each sector.21

The top panel of Table 3 shows the consequences for state/local fiscal

expenditures. Expressed as a present value at age 18, the net effect per high school

graduate is $58,990 [$15,430] and per expected graduate is $57,250 [$11,050] more

than a high school dropout’s present value. This net figure counts government

expenditures on health, crime, welfare, and taxable METB, and accounts for additional

college costs. The bottom panel of Table 3 shows significant federal fiscal expenditure

effects in addition to state/local fiscal expenditures. Federal savings per high school

graduate are $25,930 [$25,960] and $30,650 [$34,400]. Here, the fiscal consequences

are greater for state/local governments, as they are the primary funding source for

Florida’s criminal justice system and welfare system. 20 For example, the income tax “distorts” activity toward non-work: individuals would prefer to work but do not because the tax rate is too high. 21 Census (2012, p20-566.pdf); www.fldoe.org/arm/; http://controller.vpfa.fsu.edu/Student-Financial-Services/SFS-For-Students/Tuition-Rates; and Johnson (2009). Costs are inflated using the HECA index. To account for the lengthening time to degree, (two-year) four-year degrees are assumed to require (three) five years of full-time study (Hoxby and Avery, 2013, Table 1).

15


Table 4 shows the other social benefits and costs by education levels (separate

from the earnings and tax effects). College costs are higher for expected graduates, but

there are significant benefits from improved health status and reduced criminal activity.

Finally, there are productivity gains from having a more educated workforce. These

differences are substantial and apply across all groups of Florida students.

5. The Economic Burden of High School Dropout in Florida

5.1 Lifetime Burdens of High School Dropouts

The calculation of benefits by attainment level can be merged with data on the

differences in the proportions of students reaching each education level (as per Figures

1F and 1M). The total individual lifetime burden per high school dropout is given in

Tables 5 and 6.

The social benefit for the state of Florida per high school graduate is extremely

large (Table 5). Including earnings gains, improved health status and reduced

government health spending, reduced crime and reduced criminal justice system

expenditures and victim costs, and the other benefits, the social gain per graduate over

a dropout is valued at $444,380 [$263,040] per male [female] student. On average, the

social gain is $353,730 per new high school graduate. Thus, if the state could ensure

that high school dropouts did graduate from high school, this amount would be the

expected benefit. The full social benefit is even greater, because some high school

graduates enroll in postsecondary education. Even assuming a very low rate of college

enrollment, a more accurate valuation of high school graduation is on average

$475,570. A large proportion of this benefit is generated by graduates’ higher earnings,

but there are important gains to the healthcare system, the criminal justice system, and

across the Florida economy. These gains hold for both genders and racial groups.

There are also sizeable fiscal benefits when school dropout is averted. These

benefits are split between the state/local government and the federal government, even

as most educational spending is sourced by the former. Disaggregated federal benefits

are reported in Appendix Table 3. At the state/local level, the fiscal benefits per

individual are sizeable (see Table 6). From a narrow Florida taxpayer perspective, this

16


amount is what is being wasted when a student drops out instead of completing high

school. Net of college costs subsidized by the state, there are fiscal gains to healthcare

provision, the criminal justice system, and welfare programs, as well as extra tax

contributions. Per high school graduate, these gains are valued at $56,720; per

expected graduate, the gains are $62,950.

A more accurate accounting of the waste to the Florida taxpayer, however,

would include most of the consequences for the federal government. First, more than

95% of all federal government dollars that are paid in by Florida residents are

reallocated within Florida. Therefore, any federal government saving is valuable to the

state. Second, much federal government funding is subject to maintenance-of-effort

requirements (especially in the health and welfare sectors). If federal government

spending in Florida changes, there are implications for Florida state/local government

spending. The federal impacts are greater than the state impacts.

Overall, the total fiscal burden per high school dropout is $115,680 when

compared to a high school graduate or $151,560 when compared to an expected

graduate. These amounts include the cost of college but not any additional amounts

needed to ensure that high school dropouts become graduates. By contrast, the

average annual wage in Florida is $41,000 and per-capita annual general revenue for

state/local government is $6,000 (Census, 2012). By age 18, total spending per Florida

public school student from kindergarten to 12th grade will have been $167,000; the

benefits of high school graduation are almost as large as the entire amount spent on

schooling.

In aggregate, the present value burden of high school dropouts across each

cohort of Florida students is $19.1-$25.6 billion for the state. From the Florida taxpayer

perspective, the burden is $6.3-$8.2 billion. These are annual burdens, in the sense that

a sizeable group of Florida students will fail to complete high school each year. By

comparison, the gross state product of Florida is $780 billion across its 11.3 million

working-age population.22 Each year, approximately 1 percent of the state fiscal product

is being lost by the failure to ensure that all students complete high school. Gross 22 Bureau of Economic Analysis, regional data web portal, working-age population, http://edr.state.fl.us/Content/population-demographics/data/Pop_Census_Day.pdf.

17


state/local revenues are $180 annually and federal spending is $176 billion, so the fiscal

burden is valued at 3%-4% of what the state or federal government spends. It is

equivalent to or exceeds the state’s entire annual spending on policing, health, or the

criminal justice system (Appendix Table 1).

5.2 Sensitivity Analysis

These calculations of the economic value of education for Florida rely on many

datasets, research studies, and modeling relationships. These datasets are the best

available, cover recent cohorts, and, where available, are specific to the education

system and the labor market in Florida. Moreover, where possible, conservative

assumptions have been applied. It bears noting that this economic calculation looks

only at the monetized benefits of high school graduation. No account is taken of the

psychological impact of suspending students from school or the implications for families.

Nevertheless, it is necessary to see how sensitive these economic calculations are to

alternative assumptions.

When other factors are considered, it is likely that the returns to educational

attainment in Florida are substantially above those presented above. First, some

important benefits of education have not been included in these calculations. These

include the enjoyment students get from going to college and the option value for taking

an advanced degree or professional training, as well as the value of employer-provided

pension plans. The earnings profiles do not fully account for the higher probability of

being employed or accurately price out non-labor market time or account for difference

in work-life expectancies. Second, the economic model is extrapolated using current

associations. However, recent historical evidence suggests that the benefits of

education are growing—the earnings, health, and crime “education gradients” have

become steeper over recent decades—and so the returns to education will be greater in

the future. For example, the earnings gap between high school graduates and college

graduates has tripled since the 1960s (Oreopoulos and Petronijovic, 2012).

The returns are unlikely to be reduced by endogenous shifts to the labor market.

One such change would be that, if the pool of skilled workers grows, wages might be

expected to fall. However, most labor market studies find that the returns to education

18


rise when there is a larger pool of educated workers; this effect is due to skill-biased

technological change (Carnevale et al., 2010). Therefore, increasing the number of

college graduates should not decrease their wages. Another possible shift might be

driven by migration, although this too is unlikely. During the period from 2005 to 2009,

more than 500,000 people moved from Florida to other states across the U.S. (ACS 5-

Year Estimates, 2013, Table 3). Potentially, therefore, Florida might be losing many

skilled workers to other states. However, over the same period, 600,000 people moved

into Florida from other states. Thus, despites substantial year-on-year mobility, the net

effect on the state population level is small. Moreover, for migration to affect the model

results, the education levels of the in-migrants would have to be dramatically different

from the education levels of the out-migrants. In fact, among adults age 22-39, in-

migration raised the stock of human capital in Florida, as in-migrants had higher

education levels.23

To illustrate worst-case results, we perform a series of separate sensitivity tests

and recalculate the total social burden and state/local fiscal burden per student. We can

use these sensitivity tests to place lower boundaries around the returns to high school

completion.

The results for these tests, along with the baseline estimates, are given in Table

7. In the first test (S1), we apply the lowest estimates of state tax revenue from the three

methods described above. Relative to a high school graduate [expected graduate], the

state fiscal burden per dropout is estimated at $51,090 [$53,940], which is 90% [86%] of

the baseline estimate. (The social burden, which includes gross earnings not taxes, is

unchanged). The second test (S2) uses a discount rate of 10% instead of 3.5%; with

this test, the future impact (earnings gains for high school graduates) is given a lower

weight. This 10% rate is significantly above the conventional rate and above the highest

rates used in cost-benefit analysis of social investments. This assumption sharply

reduces the benefits of high school completion by more than one-half. However, the

fiscal benefits of high school completion are still non-trivial at $28,750 [27,190], and the

social benefits are still large at $156,840 [$180,530]. The third test (S3) assumes that

30% of any monetary gains associated with education are instead attributable to 23 NCHEMS Information Center, retrieved March 10, 2014, higheredinfo.org.

19


unmeasured ability or motivation. This assumption is very conservative: only a handful

of studies have found ability bias to be this high. Under this test, the social and fiscal

benefits are reduced by 10%-15%. Nevertheless, the economic benefits of high school

graduation are still strongly positive. Our final sensitivity test (S4) assumes there are

only benefits up to ten years after age 18: after that age, any economic differences are

assumed to be caused by other factors. Given that college completion typically takes 4-

6 years, this assumption is particularly strict. However, even under this worst-case

assumption there are still clear and substantial benefits from high school graduation.

The social gains of high school are $129,060 [$124,460], and the state/local fiscal

benefits are $36,250 [$37,020]. Therefore, highly restrictive assumptions still show a

large burden when students fail to complete high school.

5.3 The Economic Burden of High School Suspensions

Using the model, we can calculate the economic losses from school suspensions. We

know that school suspensions lead to higher dropout rates and that a higher dropout

rate has significant economic consequences. These consequences are shown in Table

8 for a range of new policies or practices regarding suspension.

The economic consequences for a single ninth-grade cohort are calculated as

follows. Across the student population of ninth graders in 2000, 48,850 (27%) were

suspended. We assume a similar number of suspensions in more recent cohorts. These

students are much less likely to graduate from high school. As noted above, high school

students who were suspended graduate at rates that are 6.6 percentage points lower

than students who are not suspended, even after controlling for factors that influence

both suspension rates and graduation rates (Balfanz et al., 2014).24

Thus, current school suspension policies and practices lead to 3,220 more high

school dropouts (Table 8, column 2). The social loss per dropout is $475,570, thus the 24 A slightly higher estimate is 7.3 percentage points. These estimates control for whether student was new to Florida Public Schools (mobility); Ethnicity; Overage for grade status; Free/Reduced Price Lunch program eligibility; LEP status; Special Education Status; ninth-grade Attendance rate; ninth-grade course failures; and ninth-grade FCAT Scale Score for Math (for details see Balfanz et al., 2014). Results are not separated by gender. Associations are based on the OSS and extended definition of a diploma. An alternative estimate of the association between suspension and high school graduation is calculated by Balfanz (2013): for students who are not suspended, the high school dropout rate is half that of suspended students.

20


total social loss to the state of Florida is $1.5 billion. The fiscal loss per dropout is

$151,560, thus the total fiscal loss to the state for each cohort is $488 million.

Alternatively, schools and districts might simply reduce the suspension rate. That

would have an indirect effect on the high school dropout rate. The last three columns of

Table 8 show the economic consequences when the suspension rate is reduced by 1,

5, or 10 percentage points, respectively. For example, if the suspension rate fell by 5

percentage points (from 27% to 22%) then the number of high school dropouts would

fall by 600 (Table 8, column 4). This would avert social losses of $285 million and fiscal

losses of $91 million.

6. Conclusions

Increasing educational attainment generates strong economic returns. Conservatively,

graduating from high school yields returns in excess of $350,000 for the state and

$115,000 from a fiscal perspective. These returns are spread across students,

taxpayers, and state residents, and they are robust to gender and race categorizations.

They are economically substantial when compared against median income, government

spending, or gross state product. When so many Florida students fail to graduate from

high school, the state is failing to make optimal investments in the future.

This finding holds for Florida and across the nation. In some respects, the need

for investments in Florida’s K-12 education is especially pressing. As noted above, the

state’s high school graduation is lower than the national average by at least five

percentage points, and gender and racial disparities in schooling, college, and labor

market prospects are as significant in Florida as they are in other states (Census,

2012). Moreover, the crime rate in Florida substantially exceeds the national average

(Appendix Table 2). Finally, state investment in education is relatively low: across the

U.S., state spending on education is 30% of total state government spending; in Florida,

it is 26%. The difference represents more than $1 billion annually in state spending that

is not prioritized to education (NASBO, 2012, Table 5).

Arguments that Florida is a special case—and so would not benefit from

increased spending on education—have little merit. Patterns of migration and student

mobility in Florida are not influential; given the in-migration of educated workers, recent

21


trends may even strengthen the case for investment. Even as Florida does not have a

state income tax, this does little to influence the findings. The fact that Florida relies

almost completely on (more regressive) property and sales taxes has little effect on an

economic model of high school dropouts (as opposed to very high income earners).

Across the U.S., the average state collects only one-third of its revenues from income

taxes; most rely on sales taxes for the bulk of their revenues. A final consideration is

that Florida, like all states, receives intergovernmental transfers and direct within-state

spending from the federal government. The fiscal source and incidence of the tax

therefore do not materially affect the model results.

Given the burden of high school dropouts, there is an important economic case

for investing in school practices that would ameliorate high school failure. One approach

would be to improve school suspension policies—current policies, besides varying

widely across states and with potentially disparate impact across student groups, do not

appear to be effective in helping students complete high school. If school practices

could be improved, it would most likely be the case that—net of the resources required

to effect them—these improvements would be very efficient from both fiscal and social

perspectives. Our estimates show that, for Florida, redressing the adverse impact of

suspensions on high school graduation rates would generate economic value of $1.5

billion socially and $0.5 billion from a fiscal perspective. Reducing the suspension rate—

which in turn would reduce the dropout rate—would also generate substantial savings.

Simply on the grounds of efficiency, there is a compelling case for improving school

disciplinary practices.

22


References

Abel, J.R., I. Dey, & T.M. Gabe. 2010. Productivity and the density of human capital. Working paper, Documents de Treball de l’IEB, University of Barcelona.

Adler, N.E., & J. Stewart. 2010. Health disparities across the lifespan: Meaning, methods, and mechanisms. Annals of the New York Academy of Sciences, 1186, 5-23.

Allgood, S., & A. Snow. 1998. The marginal cost of raising tax revenue and redistributing income. Journal of Political Economy, 106, 1246-1273.

Altonji, J. G., E. Blom, & C. Meghir. 2012. Human capital investments: High school curriculum, college major, and careers. Annual Review of Economics, 4(1), 185–223.

Avery, C., & S. Turner. 2012. Student loans: Do college students borrow too much—or not enough? Journal of Economic Perspectives, 26(1), 165–192.

Balfanz, R., V. Byrnes, & J. Fox. 2013. Sent home and put off-track: The Antecedents, disproportionalities, and consequences of being suspended in the ninth grade. Monograph, prepared for the Center for Civil Rights Remedies and the Research-to-Practice Collaborative.

Baum, S., J. Ma, & K. Payea. 2010. Education pays 2010: The benefits of higher education for individuals and society. College Board Advocacy and Policy Center.

Belfield, C.R., & H.M. Levin. 2007. The price we pay: The economic and social costs of inadequate education. Washington, DC: Brookings Institution.

Belfield, C.R., & T. Bailey. 2011. The returns and benefits to community college: A review of the evidence. Community College Review, 39, 46-68.

Bureau of Justice Expenditure and Employment Extracts [BJEEE]. 2008. Expenditure data for 2006. Washington, DC: BJS, NCJ224394;

Bush School of Government and Public Service. 2009. The ABCDs of Texas education: Assessing the benefits and costs of reducing the dropout rate. Texas A&M University.

Carneiro, P., J.J. Heckman, & E. Vytlacil. 2011. Estimating marginal returns to education. American Economic Review, 101, 2754-2871.

Carnevale, A.P., N. Smith, & J. Strohl. 2010. Help wanted: Projections of jobs and education requirements through 2018. Retrieved from cep.georgetown.edu/jobs2018/

Carroll, S., & E. Erkut. 2009. The benefits to taxpayers from students’ educational attainment. Santa Monica, CA: RAND Education.

Cohen, M., & A. Piquero. 2009. New evidence on the monetary value of saving a high risk youth. Journal of Quantitative Criminology, 25, 1, 25-49.

23


Cutler, D., & A. Lleras-Muney. 2010. Understanding differences in health behaviors by education. Journal of Health Economics, 29, 1-28.

Dale, S., & A.B. Krueger. 2011. Estimating the return to college selectivity over the career using administrative earnings data. NBER working paper 17159.

DeLisi, M., A. Kosloski, M. Sween, E. Hachmeister, M. Moore, & A. Drury. 2010. Murder by numbers: Monetary costs imposed by a sample of homicide offenders. Journal of Forensic Psychiatry and Psychology, 21, 501-513.

Fabelo, T., M.D. Thompson, J.D. Plotkin, D. Carmichael, M.P. Marchbanks, & E.A. Booth. 2011. Breaking schools' rules: A statewide study of how school discipline relates to students' success and juvenile justice involvement. Monograph, CSG Justice Center and PPRI.

Farrington, D., & B Welsh. 2007. Saving children from a life of crime. New York: Oxford University Press.

Chapman, C., J. Laird, N. Ifill, & A. KewalRamani. 2011. Trends in high school dropout and completion rates in the United States: 1972–2009 (NCES 2012-006). Washington, DC: U.S. Department of Education, National Center for Education Statistics.

Fletcher, J.M., & D.E. Frisvold. 2011. College selectivity and young adult health behaviors. Economics of Education Review, 30, 826-837.

Fernandes-Alcantara, A. L. 2012. Youth and the labor force: Background and trends. Monograph, Congressional Research Service, www.fas.org/sgp/crs/misc/R42519.pdf.

Gottlob, B.J. 2007. The high cost of failing to reform public education in Texas. Working paper, Washington, DC: Milton and Rose D. Friedman Foundation.

Hout, M. 2012. Social and economic returns to college education in the United States. Annual Review of Sociology, 38(August), 379-400.

Hoxby, C., & S. Turner. 2012. Expanding college opportunities for high-achieving, low-income students. SIEPR working paper, 12-014.

Johnson, N. 2009. What does a college degree cost? Delta Cost Project white paper series, http://www.deltacostproject.org/resources/pdf/johnson3-09_WP.pdf.

Kimbro, R.T., S. Bzostek, N. Goldman, & G. Rodriguez. 2008. Race, ethnicity, and the education gradient in health. Health Affairs, 27, 361-374.

Kinsler, J. 2011. School discipline: A source or salve for the achievement gap? Working paper, University of Rochester.

Knapp, L.G., J.E. Kelly-Reid, & S.A. Ginder. 2011. Enrollment in postsecondary institutions, fall 2009; graduation rates, 2003 & 2006 cohorts; and financial statistics, fiscal Year 2009 (NCES 2011-230). Washington, DC: U.S. Department of Education, National Center for Education Statistics.

24


Lochner, L., & E. Moretti. 2004. The effect of education on crime: Evidence from prison inmates, arrests, and self-reports. American Economic Review, 94, 155-189.

Lochner, L. 2011. Non-production benefits of education: Crime, health, and good citizenship. NBER working paper, w16722.

Losen, D.J., & J. Gillespie. 2012. Opportunities suspended: The disparate impact of disciplinary exclusion from school. Monograph, The Civil Rights Project, UCLA.

Losen, D.J. & T.E. Martinez. 2013. Out of school and off track: The overuse of suspensions in American middle and high schools. Monograph, The Civil Rights Project, UCLA.

Miller, T.R., M.A. Cohen, & B. Wiersema. 1996. Victim costs and consequences: A new look. National Institute of Justice Research Report, NCJ-155282.

Monaco, K., & S. Yamarik. 2013. Are there human capital externalities in U.S. states? Evidence from the Current Population Survey. Working paper, California State University, Long Beach.

Moore, M.A., A.E. Boardman, & A.R. Vining. 2013. More appropriate discounting: the rate of social time preference and the value of the social discount rate. Journal of Benefit-Cost Analysis, DOI 10.1515/jbca-2012-0008.

Muennig, P. 2007. Consequences in health status and costs. In C.R. Belfield & H.M. Levin (Eds.), The price we pay: The social and economic costs to the nation of inadequate education. Washington, DC: Brookings Institution Press.

Muennig, P., K. Fiscella, D. Tancredi, & P. Franks. 2010. The relative health burden of selected social and behavioral risk factors in the United States: Implications for policy. American Journal of Public Health, 100, 1758-1764.

Oreopoulos, P., & K.G. Salvanes. 2011. Priceless: The non-pecuniary benefits of schooling. Journal of Economic Perspectives, 25, 159-184.

Oreopoulos, P., & U. Petronijevic. 2013. Making college worth it: A review of the returns to higher education. The Future of Children, 23(1), 41–65.

Rosenblum, J. 2012. Degrees of health disparities: Health status among young adults with high school diplomas, sub-baccalaureate degrees, and baccalaureate degree. Health Services Outcome Research Methodologies, 12(2-3), 156-168.

Rotermund, S. 2008. When do California students drop out of school? CDRP Statistical Brief 9.

Rouse, C. 2007. The earnings benefits from education. In C.R. Belfield & H.M. Levin (Eds.), The price we pay: The social and economic costs to the nation of inadequate education. Washington, DC: Brookings Institution Press.

SCCJSO. 2012. Sourcebook of criminal justice statistics online. Available at www.albany.edu/sourcebook/pdf/t600032005.pdf.

25


Schoeni, R.F., W.H. Dow, W.D. Miller, & E.R. Pamuk. 2011. The economic value of improving the health of disadvantaged Americans. American Journal of Preventive Medicine, 40, S67-72.

Snyder, T. D., & S. A. Dillow. 2012. Digest of education statistics, 2011 (NCES Report No. 2012-001). Washington, DC: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics.

Sum, A., I. Khatiwade, & J. McLaughlin, with S. Palma. 2009. Key social, income, housing, civic, health and incarcerations consequences of dropping out of high school: Findings for Connecticut adults in the 21st century. Monograph, Center for Labor Market Studies, Northeastern University, www.opp.org/docs/SocialImpacts.pdf.

Trostel, P. 2010. The fiscal impacts of college attainment. Research in Higher Education, 51, 220–247.

Stillwell, R., & J. Sable. 2013. Public school graduates and dropouts from the common core of data: School year 2009–10. First look (provisional data) (NCES 2013-309rev). Washington, DC: U.S. Department of Education, National Center for Education Statistics.

U.S. Department of Health and Human Services [US DHHS]. 2012. Temporary Assistance for Needy Families program (TANF): Tenth report to Congress. Report, www.acf.hhs.gov/programs/ofa/resource/tanf-financial-data-fy-2012.

Waldfogel, J., I. Garfinkel, & B. Kelly. 2007. Public assistance programs: How much could be saved with improved education? In C.R. Belfield & H.M. Levin (Eds.), The price we pay: The social and economic costs to the nation of inadequate education. Washington, DC: Brookings Institution Press.

Wright, J.P., M.A. Morgan, M.A. Coyne, K.M. Beaver, & J.C. Barnes. 2014. Prior problem behavior accounts for the racial gap in school suspensions. Journal of Criminal Justice, 42, 257-266.

26


0

5000

10000

15000

20000

25000

White Black Hispanic Other

Figure 1F. Educa.onal A3ainment per Cohort of Florida High School Students (Female)

HS Dropout HS Graduate Some College BA Degree Plus

0

5000

10000

15000

20000

25000

White Black Hispanic Other

Figure 1M. Educa.onal A3ainment per Cohort of Florida High School Students (Male)

HS Dropout HS Graduate Some College BA Degree Plus

27


Table 1: Lifetime Individual Earnings by Education Level (Present Value at Age 18) Female Male Earnings HS

Dropout HS

Graduate Expected

HS Graduate

HS Dropout

HS Graduate

Expected HS

Graduate White $197,890 $419,120 $528,220 $430,340 $672,090 $841,450 Black $208,140 $367,100 $453,990 $313,330 $513,940 $595,170 Hispanic $207,390 $384,320 $462,580 $452,550 $604,470 $708,360 Other $222,030 $394,310 $508,300 $410,720 $713,830 $921,470 Average $206,750 $397,410 $493,960 $406,430 $623,770 $760,110

Difference over HS Dropout

+ $190,660 + $287,210

+ $217,340 + $353,680

Sources: CPS data 2009-2013; ACS data 2006-2010. Notes: Dollar amounts in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-specific and race-specific education distributions in Florida. Expected HS graduation rates weighted by sex and race. See Appendix I for adjustments for labor force participation, health benefits, ability, and productivity growth.

28


Table 2: Lifetime Individual Tax Payments by Education Level (Present Value at Age 18) Female Male Taxes HS

Dropout HS

Graduate Expected

HS Graduate

HS Dropout

HS Graduate

Expected HS

Graduate State/Local: White $55,800 $67,850 $76,620 $59,050 $76,840 $88,410 Black $43,720 $44,280 $56,350 $42,950 $61,150 $65,180 Hispanic $41,750 $61,050 $69,580 $48,570 $72,570 $82,750 Other $46,750 $73,870 $70,950 $54,560 $83,630 $99,330 Average $45,430 $61,350 $70,090 $49,700 $72,800 $82,650 Difference over HS Dropout

+ $15,920 + $24,660

+ $23,100 + $32,950

Federal: White $24,860 $55,950 $78,760 $51,780 $98,520 $134,930 Black $22,770 $43,340 $61,270 $34,570 $68,150 $83,990 Hispanic $22,160 $43,320 $58,970 $47,740 $76,370 $100,360 Other $21,030 $77,160 $85,180 $43,850 $89,210 $126,240 Average $22,790 $51,160 $70,200 $44,700 $85,990 $114,940 Difference over HS Dropout

+ $28,370 + $47,410

+ $41,290 + $70,240

Sources: CPS data 2009-2013; Florida tax code; NBER TAXSIM9. Notes: Dollar amounts rounded in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-race specific education distributions in Florida. Taxes are income tax (federal); state/county sales and property tax (state). See Appendix Box 1 for details.

29


Table 3: Lifetime Fiscal Impacts by Education Level (Present Value at Age 18) Female Male HS

Dropout HS

Graduate Expected

HS Graduate

HS Dropout

HS Graduate

Expected HS

Graduate State/Local Fiscal Expenditure: Health $20,850 $10,890 $7,920 $10,440 $5,440 $4,100 Crime $5,740 $2,230 $1,750 $71,940 $21,000 $16,840 Welfare $4,390 $3,010 $2,350 $2,610 $1,770 $1,420 College $- $- $8,330 $- $- $7,520 METB tax $1,210 $630 $790 $3,310 $1,100 $1,170 Total $32,190 $16,760 $21,140 $88,300 $29,310 $31,050 Difference over HS Dropout $15,430 $11,050 $58,990 $57,250 Federal Fiscal Expenditure: Health $33,340 $17,550 $12,770 $16,670 $8,770 $6,620 Crime $1,350 $520 $410 $16,870 $4,920 $3,950 Welfare $26,600 $18,230 $14,230 $15,840 $10,740 $8,620 College $- $- $770 $- $- $690 METB tax $2,390 $1,420 $1,100 $1,930 $950 $780 Total $63,680 $37,720 $29,280 $51,310 $25,380 $20,660 Difference over HS Dropout $25,960 $34,400 $25,930 $30,650

Sources: CPS data 2009-2013; Florida tax code. Notes: Dollar amounts rounded in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-race specific education distributions in Florida. METB tax based on earnings as per Table 1. See Appendix Box 1 for details on split between state/local and federal government in spending.

30


Table 4: Lifetime Other Social Benefits and Costs by Education Level (Present Value at Age 18) Female Male HS

Dropout HS

Graduate Expected

HS Graduate

HS Dropout

HS Graduate

Expected HS

Graduate Other Social Benefits and Costs:

College Costs $- $- $24,100 $- $- $21,740 Health gains (net) $- $16,630 $23,880 $- $16,630 $23,170 Crime burden $(12,390) $(4,810) $(3,790) $(155,420) $(45,360) $(36,380) Productivity gains $12,400 $23,850 $29,630 $24,380 $37,430 $45,610 Total $10 $35,670 $73,820 $(131,040) $8,700 $54,140 Difference over HS Dropout $35,660 $73,810 $139,740 $185,180 Sources: CPS data 2009-2013; Florida tax code; www.flboe.org. Notes: Dollar amounts rounded in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-race specific education distributions in Florida. College costs net of tuition. Health gains net of health status of HS dropouts. Crime burden includes fiscal and victim costs. Productivity gains based on earnings as per Table 1. See Appendix Box 1 for details on split between state/local and federal government in spending.

31


Table 5: Lifetime Total Social Benefits per Person by Education Level (Present Value at Age 18)


Female Male Average HS Graduate Expected

HS Graduate HS Graduate Expected


HS Graduate College Costs $- $(24,100) $- $(21,740) $- $(22,920) Earnings $190,660 $287,210 $217,350 $353,680 $204,010 $320,450 Health Savings $42,390 $57,380 $29,530 $39,560 $35,960 $48,470 Crime Savings $11,910 $13,520 $172,940 $187,050 $92,430 $100,290 Welfare Savings $1,460 $2,160 $890 $1,260 $1,180 $1,710 Productivity Gains $11,440 $17,230 $13,050 $21,220 $12,250 $19,230 METB Savings $5,180 $5,680 $10,620 $11,000 $7,900 $8,340 Total gain over HS Dropout $263,040 $359,080 $444,380 $592,030 $353,730 $475,570 Sources: Tables 1-4. Notes: Dollar amounts rounded in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-race specific education distributions in Florida.

32


Table 6: Lifetime State and Local Fiscal Savings per Person by Education Level (Present Value at Age 18)





HS Graduate State and Local Fiscal Impacts: College Costs $- $(8,330) $- $(7,520) $- $(7,930) Health Fiscal Gains $9,960 $12,930 $5,000 $6,340 $7,480 $9,640 Crime Fiscal Gains $3,510 $3,980 $50,940 $55,100 $27,220 $29,540 Welfare Fiscal Gains $1,380 $2,040 $840 $1,190 $1,110 $1,620 Tax Contributions $15,920 $24,660 $23,100 $32,940 $19,510 $28,800 METB $580 $410 $2,210 $2,150 $1,400 $1,280 Total gain over HS Dropout $31,350 $35,690 $82,090 $90,200 $56,720 $62,950 Federal tax impacts in Florida $52,690 $79,350 $65,190 $97,870 $58,960 $88,610

Total gain over HS Dropout (incl. federal tax impacts) $84,040 $115,040 $147,280 $188,070 $115,680 $151,560 Sources: Tables 1-4, Appendix Table 3. Notes: Dollar amounts rounded in present values at age 18 (d=0.035) in 2013 prices. Average and difference weighted according to sex-race specific education distributions in Florida.

33


Table 7: Sensitivity Tests: Lifetime State and Local Fiscal Savings (Present Value at Age 18) HS Graduate Expected

HS Graduate Net Gain

over HS Dropout

% of baseline

Net Gain over HS Dropout

% of baseline

Social Benefits: Baseline estimate $353,730 $475,570 S1: Lower bound for state taxes $353,730 100% $475,570 100% S2: Discount rate of 10% $156,840 44% $180,530 38% S3: Upper bound adjustment for ability $305,740 86% $399,670 84% S4: Ten-year horizon for benefits $129,060 36% $124,460 26%

State/local Government Savings: Baseline estimate $56,720 $62,950 S1: Lower bound for state taxes $51,090 90% $53,940 86% S2: Discount rate of 10% $28,750 51% $27,190 43% S3: Upper bound adjustment for ability $50,610 89% $54,330 86% S4: Ten-year horizon for benefits $36,250 64% $37,020 59% Sources: Baseline estimates from Tables 5 and 6. Lower bound for state taxes from CPS data 2009-2013; state sales tax. Ability adjustment of 30%. Ten-year horizon for incomes, taxes, and all government spending.

34


Table 8: Aggregate Burden of High School Failure and Suspension Policies for Florida

Status Quo Policy

Impact under New Policies/Practices Total Impact

of Suspensions

on HS Graduation

If Suspension Rate Reduced by:

1 pp 5 pp 10 pp

Suspension rate 27% 27% 26% 22% 17% Number of suspended students per 9th grade cohort 48,850

48,850 47,030 39,750 30,660

Reduction in number of high school dropouts

3,220 120 600 1,200

Social Loss averted ($ m)

$1,531 $57 $285 $571

Fiscal Loss averted ($ m)

$488 $18 $91 $182

Sources: Figures 1FM; Appendix Tables 1 and 2; Tables 5 and 6. Notes: PP is percentage points. Suspension effect on graduation from Balfanz et al. (2014). 2013 prices.

35


Appendices

Appendix 1 Model for Calculating the Social/Fiscal Burden of High School Failure

Burdens are calculated using research evidence on dropouts, high school graduates, college enrollees, and college completers. Differences by gender-race are capitalized over the lifecycle k=18-65 and expressed as present values at age 18 (d=3.5%). All figures are in real 2013 dollars, adjusted for Florida price levels (i). Data sources are in parentheses. Social Burden (S): S = YP + HF + HP + MF + MV + WS + YG + m(t.(YP + YG) + HF + CF + WS + T) - ES - Ep where YP Additional gross earnings HF Lower government health expenditures HP Private health gains MF Criminal Justice System expenditures MV Victim/social costs of crime WS Welfare and social service payments – non-transfer administrative costs YG Lost productivity spillovers across the workforce m Marginal Excess Tax Burden coefficient t Tax rate on income T Non-income tax amount payable ES Public support for college EP Tuition fees charged to student Fiscal Burden (F): F = t.(YP + YG) + HF + CF + WF + m(t.(YP + YG) + HF + CF + WS + T) - ES where YP Additional gross earnings HF Lower government health expenditures MF Criminal Justice System expenditures WF Welfare and social service payments YG Lost productivity spillovers across the workforce m Marginal Excess Tax Burden coefficient t Tax rate on income T Non-income tax amount payable ES Public support for college Data sources for key parameters: YP, YG : CPS, ACS, BLS, Abel et al. (2010), Carneiro et al. (2011); HF : MEPS, KFF; HP : Schoeni et al. (2011); MF : BJS, JEEES; MV : Miller et al. (1996); WS : BLS, Baum et al. (2010); m: Allgood and Snow (1998); t, T: CPS, IRS, www.taxfoundation.org, www.taxadmin.org, dor.myflorida.com/dor/taxes/sales_tax.html; ES, EP: NCES, www.deltacostproject.org, www.fldoe.org; d : Moore et al. (2013); i : CPI, HECA, BLS.

36


Appendix Table 1 Government Revenues and Expenditures in Florida

Annual Revenue and Spending

($ billions)

Percentage of Total (%)

Revenue $181.17

Intergovernmental federal revenue $32.02 18% Property tax revenue $25.93 14% Sales tax revenue $32.38 18% Other tax revenues $3.12 2% Other charges $83.89 46% Direct Expenditure $164.06 Education (all levels) $38.69 24% Public welfare $23.54 14% Hospitals $8.37 5% Police protection $7.35 4% Health $4.92 3% Correction $4.29 3% Fire protection $3.31 2% Housing and community development $3.33 2% Judicial and legal $2.20 1% Capital outlay $15.21 9% Other direct general expenditure $26.76 16% Other expenditure $26.09 16%

Source: U.S. Census Bureau, 2011 Annual Surveys of State and Local Government Finances. Table 1, tabulation date July 2013. www.nasbo.org. Notes: Other tax revenues are motor vehicle, corporate, and other. Other charges are miscellaneous general revenue, insurance trust revenue, utility revenue, and other charges. Direct expenditure is net of payments on borrowing.

37


Appendix Table 2 Criminal Activity in Florida (2012)

Number of Crimes

Florida Crime Rate

(Crimes per 100,000 persons)

Percent Florida Crime Rate

Exceeds National Average

Violent crime 94,087 487 26% Property crime 632,988 3,277 15% Aggravated assault 63,929 331 37% Burglary 153,563 795 19% Larceny-theft 442,095 2,289 17% Murder 1,009 5 11% Forcible rape 5,260 27 1% Robbery 23,889 113 -9% Motor vehicle theft 37,330 193 -16%

Source: FBI Uniform Crime Report (2012, Table 4)

38


Appendix Table 3 Lifetime Federal Fiscal Savings per Person by Education Level (Present value at age 18)





HS Graduate Federal Impacts: College costs $(770) $(690) $(730) Health fiscal gains $15,790 $20,570 $7,900 $10,050 $11,850 $15,310 Crime fiscal gains $820 $930 $11,950 $12,920 $6,390 $6,930 Welfare fiscal gains $8,370 $12,370 $5,100 $7,220 $6,740 $9,790 Tax contributions $28,370 $47,410 $41,290 $70,250 $34,830 $58,830 METB $970 $1,290.00 $970 $1,150 $970 $1,220 Total gain over HS dropout $54,320 $81,800 $67,210 $100,900 $60,780 $91,350